|Publication number||US5379414 A|
|Application number||US 07/911,409|
|Publication date||Jan 3, 1995|
|Filing date||Jul 10, 1992|
|Priority date||Jul 10, 1992|
|Publication number||07911409, 911409, US 5379414 A, US 5379414A, US-A-5379414, US5379414 A, US5379414A|
|Inventors||Phillip M. Adams|
|Original Assignee||Adams; Phillip M.|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (7), Non-Patent Citations (11), Referenced by (42), Classifications (24), Legal Events (8)|
|External Links: USPTO, USPTO Assignment, Espacenet|
1. The Field of the Invention
This invention relates to the detection and recovery procedure of an undetected Floppy Diskette Controller ("FDC" ) data error where data corruption occurs and, more particularly, to novel systems and methods implemented as a software-only device driver which eliminates the need for hardware redesign and/or fabrication of new FDCs.
2. The Background Art
Computers are now used to perform functions and maintain data which is critical to many organizations. Businesses use computers to maintain essential financial and other business data. Computers are also used by government to monitor, regulate, and even activate, national defense systems. Maintaining the integrity of the stored data is essential to the proper functioning of these computer systems, and data corruption can have serious (even life threatening) consequences.
Most of these computer systems include diskette drives for storing and receiving data on floppy diskettes. For example, an employee of a large financial institution might have a personal computer that is attached to the main system. In order to avoid processing delays on the mainframe, the employee may routinely transfer data files from the host system to his local personal computer and then back again, temporarily storing data on a local floppy diskette. Similarly, an employee with a personal computer at home may occasionally decide to take work home, transporting data away from and back to the office on a floppy diskette.
Data transfer to and from a floppy diskette is controlled by a device called a Floppy Diskette Controller ("FDC"). The FDC is responsible for interfacing the computer's Central Processing Unit ("CPU") with the physical diskette drive. Significantly, since the diskette is spinning, it is necessary for the FDC to provide data to the diskette drive at a specified data rate. Otherwise, the data will be written to the wrong location on the diskette.
The design of the FDC accounts for situations when the data rate is not adequate to support the rotating diskette. Whenever this situation occurs, the FDC aborts the operation and signals the CPU that a data underrun condition has occurred. Unfortunately, however, it has been found that a design flaw in many FDCs makes it impossible to detect all data underrun conditions. This flaw has, for example, been found in the NEC 765, INTEL 8272 and compatible Floppy Diskette Controllers. Specifically, data loss and/or data corruption can occur during data transfers to diskettes (or even tape drives and other media which employ the FDC), whenever the last data byte of a sector being transferred is delayed for more than a few microseconds. Furthermore, if the last byte of a sector write operation is delayed too long then the next (physically adjacent) sector of the diskette will be destroyed as well.
For example, it has been found that these FDCs cannot detect a data underrun on the last byte of a write operation to a sector of a diskette. Consequently, if the FDC is preempted during a data transfer (thereby delaying the transfer), and an underrun occurs on the last byte of a sector, the following occurs: (1) the underrun flag does not get set, (2) the last byte written to the diskette is made equal to the previous byte written, and (3) CRC is generated on the altered data. The result is that incorrect data is written to the diskette and validated by the FDC.
Conditions under which this problem may occur can be identified by simply identifying those conditions that can delay data transfer to the diskette drive. In general, this requires that the computer system be engaged in "multi-tasking" operation or in overlapped input/output ("I/O") operation. Multi-tasking is the ability of a computer operating system to simulate the concurrent execution of multiple tasks. Importantly, concurrent execution is only "simulated" because there is only one CPU, and it can only process one task at a time. Therefore, a system interrupt is used to rapidly switch between the multiple tasks, giving the overall appearance of concurrent execution.
MS-DOS and PC-DOS, for example, are single-task operating systems. Therefore, one could argue that the problem described above would not occur. However, there are a number of standard MS-DOS and PC-DOS operating environments that simulate multi-tasking and are susceptible to the problem. The following environments, for example, have been found to be prime candidates for data loss and/or data corruption due to the FDC: local area networks, 327x host connections, high density diskettes, control print screen operations, terminate and stay resident (TSR) programs. The problem has also been found to occur as a result of virtually any interrupt service routine. Thus, unless the MS-DOS and PC-DOS operating systems disable all interrupts during diskette transfers, they are also susceptible to data loss and/or corruption.
Perhaps the best way to demonstrate the FDC error is to simulate a great deal of system activity. In other words, make the computer system act as though it were performing a large number of complex tasks all at one time. The problem has accordingly been demonstrated in systems using MS/PC-DOS operating systems by means of a simple test program. First, a clock program is executed and becomes a TSR task having the responsibility of servicing the timer interrupt (Ox1C) and updating the time on the screen. Second, a MS/PC-DOS diskette program is executed which writes a sector to the diskette using the BIOS interface interrupt (Ox13) and then reads the sector back. Once the sector has been written and read back the data is compared to determine whether or not an undetected error has occurred. A running total of both detected and undetected errors can then be output to the display. The results of using such a test program on various machines was quite astonishing. For example, the IBM PS/2 series seemed most susceptible to the problem, with roughly a 30% undetected error rate.
The UNIX operating system is a multi-tasking operating system, and it is extremely simple to create an environment that can cause the problem. One of the more simple examples is to begin a large transfer to the diskette and place that task in the background. After the transfer has begun then begin to display (cat) the contents of a very large file. The purpose of the video access is to force the video buffer memory refresh logic on DMA channel 1, along with the video memory access, to preempt the FDC operations occurring on DMA channel 2 (which is lower priority than channel 1). This example creates the classic overlapped I/O environment and can force the FDC into an undetectable error condition. More rigorous examples could include the concurrent transfer of data to or from a network or tape drive using a high priority Direct Memory Access ("DMA") channel while the diskette transfer is active. Clearly, the number of possible error producing examples is infinite and very possible in this environment.
For all practical purposes the OS/2 operating system can be regarded as a UNIX derivative. In other words, OS/2 suffers from the same problems that UNIX does. There are, however, two significant differences between OS/2 and UNIX. First, OS/2 semaphores video updates with diskette operations in an effort to avoid forcing the FDC problem to occur. However, any direct access to the video buffer, in either real or protected mode, during a diskette transfer will bypass this safe-guard and render OS/2 in the same condition as UNIX. Second, OS/2 incorporates a unique command that attempts to avoid the FDC problem by reading back every sector that is written in order to verify that the operation completed successfully. This command is an addition to the MODE command (MODE DSKT VER=ON). With these changes, data loss and/or data corruption should occur less frequently than before, but it is still possible for the FDC problem to destroy data that is not related to the current sector operation.
There are a host of other operating systems that are susceptible to the FDC problem just like DOS, OS/2and UNIX. However, these systems may not have an install base as large as DOS, OS/2 or UNIX, and there may, therefore, be little emphasis on addressing the problem. Significantly, as long as the operating system utilizes the FDC and services system interrupts, the problem can manifest itself. This can, of course, occur in computer systems which use virtually any operating system.
Some in the computer industry have suggested that the FDC problem is extremely rare and difficult to reproduce. Admittedly, the problem is often very difficult to detect during normal operation because of its random characteristics. The only way to visibly detect this problem is to have the FDC corrupt data that is critical to the operation at hand. There may, however, be many locations on the diskette that have been corrupted, but not accessed. Studies have recently demonstrated that the FDC problem is quite easy to produce and may be more common than heretofore believed.
Computer users may, in fact, experience this problem frequently and not even known about it. After formatting a diskette, for example, the system may inform the user that the diskette is bad, although the user finds that if the operation is performed again on the same diskette everything is fine. Similarly, a copied file may be unusable, and the computer user concludes that he or she just did something wrong. For many in this high-tech world, it is very difficult to believe that the machine is in error and not ourselves. It remains a fact, however, that full diskette back-ups are seldom restored, that all instructions in programs are seldom, if ever, executed, that diskette files seldom utilize all of the allocated space, and that less complex systems are less likely to exhibit the problem.
Additionally, the first of these FDCs were shipped over 10 years ago. The devices were primarily used at that time in special-purpose operations in which the FDC problem would not normally be manifest. Today, on the other hand, the FDCs are incorporated into general-purpose computer systems that are capable of concurrent operation (multi-tasking or overlapped I/O). Thus, it is within today's environments that the problem is most likely to occur by having one of the operations delay the data transfer to the diskette. The more complex the computer system, the more likely it is to have one activity delay another, thereby creating the FDC error condition.
In short, the potential for data loss and/or data corruption is present in all computer systems that utilize this type of FDC, presently estimated at about 25 million personal computers. The design flaw in the FDC causes data corruption to occur and manifest itself in the same manner as a destructive computer virus. Furthermore, because of its nature, this problem has the potential of rendering even secure databases absolutely useless.
Those skilled in the art have suggested various ways of addressing the FDC problem. Unfortunately, however, each of these prior solutions has significant associated costs, risks and/or disadvantages.
For example, perhaps the most desirable solution is to have the manufacturer of the FDC provide a new FDC that alleviates the problem. This approach is, however, only a partial solution since many of the current systems have the FDC soldered into a circuit board. It would, of course, entail significant effort and/or cost to remove the current FDC and replace it with a new one.
Add-on hardware devices have similarly been suggested which could detect the FDC error condition and force it to be acknowledged by the CPU. Like a new FDC, however, such devices are at best inconvenient to install and use and are thus unlikely to be used by many computer users.
In an effort to avoid the disadvantages of a hardware solution, some read back and verify programs, like the IBM OS/2 MODE command, have been developed and installed. Such programs typically require that the FDC device driver perform single-sector writes, read the previously written sector back into a sector buffer in the FDC device driver, and then compare the data that was supposed to be written to the floppy with the data contained in the readback buffer. This process is performed until all data compares properly.
There are a number of problems that occur when employing this detection and a recovery procedure. Three of the most important problems are: (1) the size of the FDC device driver grows due to sector readback buffers required; (2) unacceptable performance is encountered because each sector must be written, the diskette must then make a full revolution for the sector to be readback, and finally the readback buffer must be compared with the original data to determine the success or failure of the I/O operation (thus causing all diskette transfers to execute at roughly one-third their normal speed); and (3) this approach is only partially effective in eliminating the FDC problem since it does not account for the data corruption that can occur to the physically adjacent sector when data transfer is significantly delayed. In short, the write/read/compare approach does not adequately protect the data from being corrupted, it causes more memory to be utilized by the operating system, and it degrades performance of the floppy diskette to an intolerable level. As a result, this approach has likewise not generally been adopted.
In view of the foregoing, it is a primary object of the present invention to provide a system and method for the detection and recovery procedure of an undetected FDC data error where data corruption occurs.
It is also an object of the present invention to provide a software-only device driver which eliminates the need for hardware redesign and/or fabrication of new FDCs.
In addition, it is an object of the present invention to provide a solution to an I/O controller's (FDC) defect using DMA shadowing.
It is a further object of the present invention to provide a system and method for the detection and recovery procedure of an undetected FDC data error which reduces system performance only a minimal amount during floppy write operations.
It is a still further object of the present invention to provide a solution to an I/O controller's (FDC) defect using variable processor speed DMA shadowing via a software decoding network.
Consistent with the foregoing objects, and in accordance with the invention as embodied and broadly described herein, a system and method are disclosed in one embodiment of the present invention as including a device driver that is capable of detecting an undetectable data corruption problem without hardware redesign and/or internal modification to an existing FDC. The approach taken consists of DMA shadowing and use of a software decoding network which allows the implementation of the invention to require a small amount of memory and only degrade the performance of the computer system a minimal amount when floppy diskette write operations occur.
The foregoing and other objects and features of the present invention will become more fully apparent from the following description and appended claims, taken in conjunction with the accompanying drawings. Understanding that these drawings depict only typical embodiments of the invention and are, therefore, not to be considered limiting of its scope, the invention will be described with additional specificity and detail through use of the accompanying drawings in which:
FIG. 1 is simplified block diagram which illustrates the architecture of most computer systems employing an floppy diskette controller (FDC);
FIG. 2 is a block diagram which illustrates the typical association between application programs, operating systems, device drivers and computer system hardware (in this example, a floppy diskette);
FIG. 3 is a flow chart depicting one presently preferred embodiment of the modifications that are applied to the diskette device driver in order to allow the error detection/prevention system and method of the present invention to be activated;
FIG. 4 is a flow chart depicting one presently preferred embodiment of the modifications that are made to the timer Interrupt Service Routine (ISR) so as to allow timing of the last byte's DREQ/DACK cycle in accordance with the present invention; and
FIG. 5 is a flow chart depicting one presently preferred embodiment of a software decoding network (software vector-table) for use in connection with the error detection/prevention system and method of the present invention, the network having one code point/entry for each possible transfer byte in the sector.
It will be readily understood that the components of the present invention, as generally described and illustrated in the Figures herein, could be arranged and designed in a wide variety of different configurations. Thus, the following more detailed description of the embodiments of the system and method of the present invention, as represented in FIGS. 1 through 5, is not intended to limit the scope of the invention, as claimed, but it is merely representative of the presently preferred embodiments of the invention.
The presently preferred embodiments of the invention will be best understood by reference to the drawings, wherein like parts are designated by like numerals throughout.
The architecture of a typical computer system is illustrated in FIG. 1. The computer system's Central Processing Unit ("CPU") 10 and main memory 14 are located inside the system unit. The instructions and data used by the CPU 10 are kept in main memory 14 during computer work sessions. Main memory 14 is, however, not a permanent storage place for information; it is active only when the computer system is on. Thus, to avoid losing the data, it must be saved on some type of storage device. For example, the computer system may use a "hard disk" storage device which is permanently installed in the computer system. Most computer systems have at least one floppy diskette drive 50 that receives a removable floppy diskette. The floppy diskette likewise is used for "permanent" storage of data or software outside of the computer system and is especially useful for transferring data and information between separate computer systems.
In transferring data to a floppy diskette, the CPU 10 typically programs the Direct Memory Access ("DMA") controller 30 for an input/output ("I/O") transfer, issues a command to the Floppy Diskette Controller ("FDC") 20 to begin the I/O transfer, and then waits for the FDC to interrupt with a completion interrupt signal. It is also possible to perform Programmed I/O ("PIO") directly between the CPU 10 and the FDC 20 without involving the DMA controller 30, as illustrated by the broken arrows 16. This latter approach is seldom used; the majority of computer systems employ DMA I/O transfers to and from the floppy diskette. The invention will thus be described below with particular reference to the DMA controller 30. If PIO is employed, however, then the I/O transfer is totally controlled by the CPU 10 because the CPU is required to pass each and every data byte to the FDC 20. As a result, the "DMA shadowing" system and method of this invention can be directly applied to the PIO data stream because the CPU 10 already is controlling the I/O transfer, as will become more readily apparent from the discussion which follows.
Virtually all computer systems must have a system clock 12. The system clock is necessary when initiating an I/O transfer to the diskette drive 50 because one must not only control the data transfer, but also the drive motor. In this regard, it is important to know when the diskette drive motor has brought the diskette spin rate up to the nominal RPM required for the data transfer to be successful.
As an example, in IBM Personal Computers and compatibles, the system clock 12 interrupts the CPU 10 18.6 times per second (roughly once every 54 milliseconds). This interrupt is used to determine such things as diskette drive motor start and stop time. There are also a host of other time-dependent operations in the computer system that require this granularity of timing.
The typical association between application programs, operating systems, device drivers and hardware is depicted in FIG. 2. The example presented is the floppy diskette.
As illustrated in FIGS. 3 and 4, the system and method of the present invention includes an interposer routine 70 which is placed between the application's request for floppy service and the current floppy device driver. The interposer routine 70 is actually a new or modified device driver that implements the detection and recovery for the undetected FDC data corruption. As shown, the interposer 70 first determines if the operation is a floppy diskette write operation (72). If so, the major function of interposer 70 is to insert itself between the application request for floppy service and the floppy device driver that will service the request. In the PC/MS-DOS environment, this can be accomplished by "hooking" the INT Ox13 interrupt vector (74) and directing it to the interposer routine and then reprogramming the timer to interrupt faster than normal (76) (e.g., every 4-7 milliseconds).
As will become more fully apparent from the discussion which follows, once a floppy write operation is detected, a software decoding network call vector is preferably installed (see FIG. 5), the current byte count is read, and DMA shadowing begins. If the time is too long, an error condition is forced (86). The system clock 12 is then reprogrammed (82) to interrupt normally (e.g., every 54 milliseconds), and timer interrupt is "unhooked" (84) until the next floppy write operation. Clearly, one could allow the timer to always interrupt at the accelerated rate and then check in the timer Interrupt Service Routine ("ISR") if a diskette write operation is active, but this approach is not as performance-minded as the one presented herein.
As used herein, "DMA shadowing" means monitoring byte transfers and then timing the last byte of a sector's DREQ to DACK signals. Importantly, there are, of course, a number of ways of determining when the DREQ is present and when the DACK is present. The sample code set forth below is only one approach. The present invention includes the use of any "DMA shadowing" whether the DREQ and DACK signals are detected at the DMA controller, CPU, system bus or FDC. This includes both explicit means, as presented in the sample code, and implicit means, such as inferring the state of the DREQ/DACK cycle from various components in the system that are triggered or reset from such signal transitions. A typical example is that the DACK can cause the Terminal Count (TC) signal to be asserted. Therefore, one can imply from the detection of TC that DACK has occurred.
In other words, whenever an application requests a floppy write operation, the system clock 12 is reprogrammed to interrupt every 4 to 7 milliseconds. Each time the system clock interrupts, the current byte count in the DMA controller transfer register (countdown register) is read. Once the byte counter has reached the last byte, the signal transition from DREQ to DACK is timed. If the time is greater than the time that will insure data integrity, an error condition is forced which is similar to the one the FDC hardware would produce if not defective. Finally, to avoid corruption to the adjacent sector the system causes the FDC to abort an operation if the timing between the DREQ and the DACK extends past the corruption point, but not to the adjacent sector destruction point, and return an error condition as previously described.
In order for the system to maintain proper operation, it is necessary that the interposer 70 save the original INT Ox13 contents (address of the original INT Ox13 Interrupt Service Routine) and invoke the original when necessary. Additional aspects of the interposer function are discussed below in connection with the other features of the device driver.
The following code fragment from the actual device driver demonstrates the interposer operation. This code fragment, together with the other sample code set forth herein, illustrates in more detail one preferred embodiment of a software routine derived from the block diagram of FIGS. 3 through 5. Those of ordinary skill in the art will, of course, appreciate that various modifications to the specific sample code may easily be made without departing from the essential characteristics of the invention. Thus, the illustrative code, and the accompanying description, is intended only as an example, and it simply illustrates one presently preferred embodiment that is consistent with the invention as claimed herein.
The examples set forth below relate to an implementation of the system and method of the present invention for use on an IBM Personal Computer running the PC/MS-DOS operating system. Similar versions have, however, been developed to operate in the UNIX and OS/2 environments. The invention is not limited to use with any particular operating system, and adaptations and changes which may be required for use with other operating systems will be readily apparent to those of ordinary skill in the art.
__________________________________________________________________________PROCEDURE: .sub.-- INT13.sub.-- isrREMARKS: .sub.-- INT13.sub.-- isr is responsible for receiving the INT 13 diskette interrupts from the O/S (BIOS). A check is made to see if the requested operation is a WRITE and the drive is a diskette (0 or 1). If so, then the timer is enabled and the system is set-up to delay the last (512th) byte of the transfer to generate an undetected underrun/overrun condition..sub.-- INT13.sub.-- isr proc farmov .sub.-- TEXT : active, 0 ; Clear Diskette Write Active Flagmov .sub.-- TEXT : errors, 0 ; Clear Diskette Error Flagcmp dl,2 ; Check Drive Number for Diskette 0 or 1jb chk.sub.-- write ; If Diskette, Then Check for Writejmp .sub.-- TEXT : int13.sub.-- ptr ; Otherwise, enter Original INT 13chk.sub.-- write:cmp ah,3 ; Check For Diskette Write Operationje set.sub.-- active ; If Write, Then Diskette Write Activejmp .sub.-- TEXT : int13.sub.-- ptr ; Otherwise, Enter Original INT 13set.sub.-- active:push ds ; Save DS Registermov .sub.-- TEXT : sectors,al ; Save Number of Sectors to Transfermov .sub.-- TEXT : track,ch ; Save Track Numbermov .sub.-- TEXT : sector,cl ; Save Sector Numbermov .sub.-- TEXT : head, dh ; Save Head Numbermov .sub.-- TEXT : drive,dl ; Save Drive Numbermov .sub.-- TEXT : active,OFFH ; Set Diskette Write Active Flagcall NEAR PTR.sub.-- Timer.sub.-- enable ; Enable Timer for "Shadowing" DMApop ds ; Restore DS Registerpushf ; Simulate An Interruptcall .sub.-- TEXT : int13.sub.-- ptr ; Enter Original INT 13 ISRpushf ; Save INT 13 Flagscall NEAR PTR.sub.-- Timer.sub.-- disable ; Disable the Timer "Shadow" Routinemov .sub.-- TEXT : active,0 ; Clear Diskette Write Active Flagpopf ; Restore INT 13 Flagscmp .sub.-- TEXT : errors,0 ; Were Errors Detected by Timer ISRjz .sub.-- INT13.sub.-- exit ; If Zero, Then no Errors-Exitstc ; Otherwise, Set Error Indicatormov .sub.-- TEXT : errors,0 ; And Clear Error Flag.sub.-- INT13.sub.-- exit:ret 2 ; Eliminate Entry Flags.sub.-- INT13.sub.-- isr endp__________________________________________________________________________
The foregoing interposer routine checks to see if the request is a write operation. If so, then it calls - Timer- enable (reprogram the system clock), calls the original INT Ox13 Interrupt Service Routine (perform the actual write operation while DMA Shadowing is enabled), and finally calls -- Timer-- disable (reprograms the system clock to the original clock interrupt rate of approximately 54 milliseconds).
As mentioned above, the interposer invokes system clock management routines (-- Timer-- enable and -- Timer-- disable). These two functions provide the system clock to interrupt at an accelerated rate that is 8 to 10 times faster than normal. The disable routine (see FIG. 3) returns the system clock interrupt rate to the normal interrupt rate. Thus, they are inverse identities which make the system behave normally at all time without significant performance degradation (<10%) and only during floppy write operations. At any other time the performance is virtually unchanged.
The following code fragment depicts the operation of the -- Timer-- enable function (76):
__________________________________________________________________________PROCEDURE: .sub.-- TIMER.sub.-- enableREMARKS: Timer enable is responsible for 1) moving INT 8 vector to INT 0 × 60, 2) Initializing INT 8 vector to .sub.-- Timer.sub.-- irs, and 3) Reprogramming the 8253 to interrupt 128 times faster. (Clock count * 0.840 Microseconds).sub.-- Timer.sub.-- enable proc nearpushf ; Save Current Flags Registercli ; Disable Interruptspush ax ; Save AX Registerpush bx ; Save BX Registerpush es ; Save ES Registerxor ax,ax ; Zero AX Registermov es,ax ; Set ES to Absolute Zero Segmentmov ax,es:[8*4] ; Obtain Offset of INT 8mov bx,es:[(B*4)+2] ; Obtain Segment of INT 8mov .sub.-- TEXT : int8.sub.-- off,ax ; Move INT 8 Offset to INT 8 Offsetmov .sub.-- TEXT : int8.sub.-- seg,bx ; Move INT 8 Segment to INT 8 Segmentmov ax,es:[1CH*4] ; Obtain Offset of INT 1Cmov bx,es:[(1CH*4)+2] ; Obtain Segment of INT 1Cmov .sub.-- TEXT : int1C.sub.-- off,ax ; Move INT 1C Offset to INT IC Offsetmov .sub.-- TEXT : int1C.sub.-- seg,bx ; Move INT 1C Segment to INT 1C Segmentmov .sub.-- TEXT : ticks,0 ; Zero ticks for Use by New ISRin al,020H ; Obtain 8259 Interrupt Mask Registerpush ax ; Save 8259 IMR for Restorationmov al,OFFH ; Mask All External Interrupts At 8out 021H,al ; Set-up New 8259 IMRmov al,36H ; Establish Operational Mode of 8253out 43H,al ; Program Operation of 8253jmp $+2 ; Allow Time for I/O to Complete Opmov al,0 ; Load LSB of Down Counterout 40H,al ; Send to 8253jmp $+2 ; Allow Time for I/O to Complete Opmov al,02H ; Load MSB of Down Counterout 40H,al ; Send to 8253jmp $+2 ; Allow Time for I/O to Complete Opmov ax,OFFSET .sub.-- TEXT : .sub.-- Timer.sub.-- isr ; Obtain the Offset of .sub.-- Timer .sub.-- isrmov bx,cs ; Obtain the Segment of .sub.-- Timer.sub.-- isrmov es:[8*4], ax ; Install new INT 8 Offsetmov es:[(8*4)+2],bx ; Install new INT 8 Segmentmov ax,OFFSET.sub.-- TEXT : .sub.-- INT1C.sub.-- isr ; Obtain the Offset of .sub.-- INT.sub.-- 1Cmov bx,cs ; Obtain the Segment of .sub.-- INT.sub.-- 1Cmov es: [1CH*4]ax ; Install new INT 1C Offset; mov es:[(1CH*4)+2],bx ; Install new INT 1C Segment; pop ax ; Obtain old 8259 IMR from Stack; out 021H,al ; Set-up old 8259 IMREnable.sub.-- exit:pop es ; Restore ES Registerpop bx ; Restore BX Registerpop ax ; Restore AX Registerpopf ; Restore Flags Registerret ; Return to Caller.sub.-- Timer.sub.-- enable endp__________________________________________________________________________
The foregoing -- Timer-- enable routine performs two major functions. First, this function "hooks" the interrupt vectors that are associated with system clock functions (INT Ox08 and INT Ox1C). The original values of these two interrupt vectors are saved for use in the newly installed interrupt service routines for these interrupts (-- Timer-- isr and -- INT1C-- isr). Next, the system cloak is reprogrammed to interrupt at the accelerated rate. However, now the system clock interrupts will be processed by the newly installed system clock interrupt service routines.
The following code fragment depicts the operation of the -- Timer-disable function (82):
__________________________________________________________________________PROCEDURE: .sub.-- Timer.sub.-- disableREMARKS: .sub.-- Timer.sub.-- disable is responsible for 1) Reprogramming the 8253 to interrupt 128 times slower and 2) Restoring the old INT 8 interrupt vector from INT 60..sub.-- Timer.sub.-- disable proc near pushf ; Save Current Flags Register cli ; Disable Interrupts push ax ; Save AX Register push bx ; Save BX Register push es ; Save ES Register; in al,020H ; Obtain 8259 Interrupt Mask Register; push ax ; Save 8259 IMR For Restoration; mov al,OFFH ; Mask All External Interrupts At 8; out 021H,al ; Set-up new 8259 IMR mov al,36H ; Establish Operational Mode of 8253 out 43H,al ; Program Operation of 8253 jmp $+2 ; Allow Time for I/O to Complete Op mov al,0 ; Load LSB of Down Counter out 40H,al ; Send LSB to 8253 jmp $+2 ; Allow time for I/O to Complete Op out 40H,al ; Send MSB to 8253 jmp $+2 ; Allow time for I/O to Complete Op xor ax,ax ; Zero AX Register mov es,ax ; Set ES to Absolute Zero Segment mov ax,.sub.-- TEXT : int8.sub.-- off ; Obtain Offset of Original INT 8 mov bx,.sub.-- TEXT : int8.sub.-- seg ; Obtain Segment of Original INT 8 mov es:[8*4],ax ; Install Old INT 8 Offset mov es:[(8*4) +2], bx ; Install Old INT 8 Segment mov ax,.sub.-- TEXT : int1C.sub.-- off ; Obtain Offset of Original INT 1C mov bx,.sub.-- TEXT : int1C.sub.-- seg ; Obtain Segment of Original INT 1C mov es:[1CH*4],ax ; Install Old INT 1C Offset mov es:[(1CH*4)+2],bx ; Install Old INT 1C Segment; pop ax ; Obtain Old 8259 IMR From Stack; out 021H,al ; Set-up Old 8259 IMR pop es ; Restore ES Register pop bx ; Restore BX Register pop ax ; Restore AX Register popf ; Restore Flags Register ret ; Return to Caller.sub.-- Timer.sub.-- disable endp__________________________________________________________________________
The -- Timer-disable function above is responsible for restoring the system to its original state. In other words, the INT Ox08 and INT Ox1C interrupts are restored to their original values and the system clock is reprogrammed to interrupt at its original frequency.
As depicted graphically in FIG. 4 and described further below, a -- Timer-isr routine is used for servicing the accelerated interrupt rate of the system clock. The reason that the system clock interrupt rate is accelerated is that during a normal 512 byte data transfer (the typical sector size) 16 microseconds are required for each data byte to be transferred to the FDC (High Density Diskette Mode). Therefore, a typical Sector transfer requires 512 times 16 microseconds, or 8,192 microseconds. If the diskette is a low density diskette then the sector transfer time is doubled to 16,384 microseconds (512 times 32 microseconds) because the FDC has half of the amount of data to store in the same rotational time frame (typically 360 RPM).
The following code fragment illustrates the -- Timer-- isr routine:
__________________________________________________________________________PROCEDURE: .sub.-- Timer.sub.-- isrREMARKS: .sub.-- Timer.sub.-- isr is responsible for receiving the INT 8 timer interrupts from the 8253. Since the 8253 has been programmed to interrupt approximately 128 times faster .sub.-- Timer.sub.-- isr must call the old INT 8 ISR routine every 128th interrupt. .sub.-- Timer.sub.-- isr is also responsible for initiating a DMA request on every timer interrupt..sub.-- Timer.sub.-- isr procinc .sub.-- TEXT : ticks ; Increment the Interrupt Tick Countercmp .sub.-- TEXT : active, OFFH ; Check to see if Diskette Write Opjz begin.sub.-- isr ; If so, Then Begin Timer ISRjmp chk.sub.-- old ; Otherwise, Check if Old ISR Runsbegin.sub.-- isr:cld ; Clear the Direction Flagmov .sub.-- TEXT : AX.sub.-- reg,ax ; Save AX Registermov .sub.-- TEXT : BX.sub.-- reg,bx ; Save BX Registermov .sub.-- TEXT : CX.sub.-- reg,cx ; Save CX Registermov .sub.-- TEXT : DX.sub.-- reg,dx ; Save DX Registermov .sub.-- TEXT : SI.sub.-- reg,si ; Save SI Registerout OCH,al ; Clear the Byte Pointer Flip/Flopcmp .sub.-- TEXT : CHN1.sub.-- flag,0 ; Should Channel 1 Be Masked (Diabl)jz bypass.sub.-- CHN1 ; If Zero, Then Don't Mask Channelmask.sub.-- CHN1:mov al,005H ; DMA Mask Register Valueout 00AH,al ; Disable DMA Channels (1 for Now)jmp $+2 ; (Can't Disable Refresh on XT - CHbypass.sub.-- CHN1:; in al,020H ; Obtain 8259 Interrupt Mask Register; push ax ; Save 8259 IMR for Restoration; mov al,OBDH ; Disable all but KBD and Diskette; out 021H,al ; Set-up New 8259 IMR in al,08H ; Obtain 8237 DMA Status test al,04H ; Check for Channel 2 DMA TC je chk.sub.-- change ; If no TC, Then Check for DMA Req jmp exiting ; Otherwise, Exit Routinechk.sub.-- change:in al,05Hmov dl,alin al,05Hmov dh,almov cx,.sub.-- TEXT : ratiochk.sub.-- change.sub.-- A:in al,05Hmov bl,alin al,05Hmov bh,alcmp bx,dxjne chk.sub.-- countin al,05Hmov bl,alin al,05Hmov bh,alcmp bx,dxjne chk.sub.-- countloop chk.sub.-- change.sub.-- A ; Loop if Processor is Fastjmp exiting ; If not DRQ, Exit Routinechk.sub.-- count:mov si,bx ; Save Current DMA Countand si,01FFH ; Isolate Lower Sector Countshl si,1 ; Word Normalize SI Valuejmp .sub.-- TEXT : DMA.sub.-- Count.sub.-- TBL[SI] ; Shadow the DMA Transfers Via__________________________________________________________________________ S/W
The foregoing -- Timer-- isr routine performs some sanity checks on the system to determine if the system is actually transferring data to the FDC. If a sector transfer is not in progress then -- Timer-- isr exits immediately. However, if a sector transfer is in progress then -- Timer-- isr obtains the remaining byte count of the sector transfer and vectors (jumps) through the software decoding network (DMA-- Count-- TBL) to the appropriate processing routine.
Although the system and method depicted in FIG. 4, could be implemented as is, it would require the timer to interrupt every 8, 16, or 32 microseconds. This level of interrupts would totally consume a PC's processing power, and on a PC/XT would/could not be sustained. Thus, in order to perform DMA shadowing without affecting the total system performance it is necessary to allow normal operations to continue as usual, but have an interrupt (the system clock) that will interrupt close to the end of the sector transfer so that the DREQ to DACK timing can be determined on the last byte of the sector transfer.
Clearly, it is possible to DMA shadow all 512 bytes during a sector transfer, but that would cause the CPU to be totally consumed during the entire sector transfer time. In other words, the potential of losing processing activities somewhere else in the system are greatly increased, such as in the case of serial communications. Therefore, the following clock interrupt strategy was developed to reduce the CPU involvement to a bare minimum during the floppy write operations with DMA Shadowing. Significantly, the timing strategy can be adjusted to any number of bytes of the sector transfer, from a few bytes to the entire sector count.
As indicated above, the last operation performed in the -- Timer-- isr routine is to vector through the software decoding network to the appropriate processing routine. This process is illustrated graphically in FIG. 5. The software decoding network (software vector-table) has one code point/entry for each possible transfer byte in the sector. The timer interrupt rate can now be in terms of 10's or 100's of byte transfer times because the vector table will cause the program execution to enter a cascade of DREQ/DACK checks only when it knows that the transfer (sector) will be complete prior to another timer interrupt. In short, the first X entries in the vector table will simply return (knowing that another timer interrupt will occur before the sector transfer completes). The latter Y entries will cascade from one DREQ/DACK detection to another (shadowing the DMA transfers) until the last byte is transferred. On the last byte being transferred, an instruction loop is counted and finally converted into microseconds to determine if the error has occurred.
This process is the fastest known technique for decoding and executing time-dependent situations. Memory space (the software decoding network vector table) is traded for processing time (the amount of time it would take for one routine to subsume all functionality encoded in each of the routines vectored to through the software decoding network vector table).
The following code fragment depicts a portion of the software decoding network employed in the DMA shadowing process of the presently preferred embodiment of the system and method of the present invention:
__________________________________________________________________________DMA.sub.-- Count.sub.-- TBLLABLE WORDDW OFFSET DMA.sub.-- 0 ; Entry for DMA Count 0DW OFFSET DMA.sub.-- 1 ; Entry for DMA Count 1DW OFFSET DMA.sub.-- 2 ; Entry for DMA Count 2DW OFFSET DMA.sub.-- 3 ; Entry for DMA Count 3DW OFFSET DMA.sub.-- 4 ; Entry for DMA Count 4DW OFFSET DMA.sub.-- 5 ; Entry for DMA Count 5DW OFFSET DMA.sub.-- 6 ; Entry for DMA Count 6DW OFFSET DMA.sub.-- 7 ; Entry for DMA Count 7DW OFFSET DMA.sub.-- 8 ; Entry for DMA Count 8DW OFFSET DMA.sub.-- 9 ; Entry for DMA Count 9DW OFFSET DMA.sub.-- a ; Entry for DMA Count aDW OFFSET DMA.sub.-- b ; Entry for DMA Count bDW OFFSET DMA.sub.-- c ; Entry for DMA Count cDW OFFSET DMA.sub.-- d ; Entry for DMA Count dDW OFFSET DMA.sub.-- e ; Entry for DMA Count eDW OFFSET DMA.sub.-- f ; Entry for DMA Count fDW OFFSET DMA.sub.-- 10 ; Entry for DMA Count 10DW OFFSET DMA.sub.-- 11 ; Entry for DMA Count 11DW OFFSET DMA.sub.-- 12 ; Entry for DMA Count 12DW OFFSET DMA.sub.-- 13 ; Entry for DMA Count 13DW OFFSET DMA.sub.-- 14 ; Entry for DMA Count 14DW OFFSET DMA.sub.-- 15 ; Entry for DMA Count 15DW OFFSET DMA.sub.-- 16 ; Entry for DMA Count 16DW OFFSET DMA.sub.-- 17 ; Entry for DMA Count 17DW OFFSET DMA.sub.-- 18 ; Entry for DMA Count 18DW OFFSET DMA.sub.-- 19 ; Entry for DMA Count 19DW OFFSET DMA.sub.-- 1a ; Entry for DMA Count 1aDW OFFSET DMA.sub.-- 1b ; Entry for DMA Count 1bDW OFFSET DMA.sub.-- 1c ; Entry for DMA Count 1cDW OFFSET DMA.sub.-- 1d ; Entry for DMA Count 1dDW OFFSET DMA.sub.-- 1e ; Entry for DMA Count 1eDW OFFSET DMA.sub.-- 1f ; Entry for DMA Count 1fDW OFFSET exiting ; Address Of Exit CodeDW OFFSET exiting ; Address Of Exit CodeDW OFFSET exiting ; Address Of Exit CodeDW OFFSET exiting ; Address Of Exit CodeDW OFFSET exiting ; Address Of Exit CodeDW OFFSET exiting ; Address Of Exit CodeDW OFFSET exiting ; Address Of Exit CodeDW OFFSET exiting ; Address Of Exit CodeDW OFFSET exiting ; Address Of Exit CodeDW OFFSET exiting ; Address Of Exit CodeDW OFFSET exiting ; Address Of Exit CodeDW OFFSET exiting ; Address Of Exit CodeDW OFFSET exiting ; Address Of Exit CodeDW OFFSET exiting ; Address Of Exit CodeDW OFFSET exiting ; Address Of Exit CodeDW OFFSET exiting ; Address Of Exit Code__________________________________________________________________________
As indicated above, the entire software decoding network table is initially set to the address of the "exiting routine." Then depending upon how slow or fast the system clock interrupts a certain number of the lower-indexed entries of the table are set to the address of a processing routine. These processing routines are identical and are sequentially located in the routine so that the software decoding network vector table simply vectors the timer interrupt routine to the first of n sequentially executed processing routines where n represents the number of bytes remaining in the sector transfer. In this way the last few bytes of the sector transfer can be accurately monitored (DMA Shadowing) without significantly affecting overall system performance.
Each of the processing routines, except the last one, performs exactly the same function as indicated below:
______________________________________DMA.sub.-- 1f: mov dx,bx mov cx,5DMA.sub.-- 1f.sub.-- A: in al,05H mov bl,al in al,05H mov bh,al cmp bx,dx jne DMA.sub.-- 1e in al,05H mov bl,al in al,05H mov bh,al cmp bx,dx jne DMA.sub.-- 1e loop DMA.sub.-- 1f.sub.-- ADMA.sub.-- 1e: mov dx,bx mov cx,5DMA.sub.-- 1e.sub.-- A: in al,05H mov bl,al in al,05H mov bh,al cmp bx,dx jne DMA.sub.-- 1d in al,05H mov bl,al in al,05H mov bh,al cmp bx,dx jne DMA.sub.-- 1d loop DMA.sub.-- 1e.sub.-- A . . .______________________________________
The above routines represent the code required to completely monitor (shadow) the DMA process by watching the DREQ and DACK signal through the DMA controller. In other words, by watching the DMA controller's register that indicates when a DMA request is active (DREQ) then it is possible to completely monitor the sector transfer. It is not necessary to concern ourselves with the timing between the DREQ and DACK signals until the very last data byte of the transfer. Therefore, the routines above simply "shadow" the DMA process until the last byte at which time it is necessary to invoke the error detection and recovery procedure. When the DMA controller transfer register contains a value of 16 or less then the CPU begins to execute inline code that watches each data byte transfer in terms of signal requests and acknowledgements.
Once the last byte to be transferred has been identified then the DMA-- O routine begins the process of determining the actual time taken between the DMA Request (DREQ) and the FDC's DMA Acknowledgement (DACK). This process is presented below:
__________________________________________________________________________DMA.sub.-- 0 mov cx,.sub.-- TEXT : ratio ; Up to 32 usec Before DREQ shl cx,1 shl cx,1DMA.sub.-- DRQ.sub.-- LO: in al,08H ; Read DMA Status Register test al,40H ; Is DRQ Active? jne DMA.sub.-- DRQ.sub.-- HI ; If Non-Zero, Then DRQ is Active loop DMA.sub.-- DRQ.sub.-- LO ; Otherwise, Continue Checking jmp exiting ; DRQ Not Active so ExitDMA.sub.-- DRQ.sub.-- HI: mov cx,OFFFFH ; Load Maximum Loop CountDMA.sub.-- DRQ.sub.-- ACTIVE: in al,08H ; Read DMA Status Register test al,04H ; Is DRQ Active? je DMA.sub.-- DRQ.sub.-- INACTIVE ; If Zero, Then DRQ is Inactive loop DMA.sub.-- DRQ.sub.-- ACTIVE ; Otherwise, Continues Active jmp exiting ; DRQ Stuck High (Terrible Error)DMA.sub.-- DRQ.sub.-- INACTIVE: cmp cx,OFFFFH ; Check to See if DREQ Remained Active jne DMA.sub.-- TIME.sub.-- ACTIVE ; If so, Then Compute Time Active jmp exiting ; Otherwise, Exit NormallyDMA.sub.-- TIME.sub.-- ACTIVE: mov ax,.sub.-- TEXT : avg ; Load AX With Average 1 Loop Time not cx ; Convert Count Down Value To ; Times Through The Active Loop ; (Value Used for "n" in Equation) inc cx ; Account for First Time Through Loop mul cl ; Compute (n * AVG) mov .sub.-- TEXT : time.sub.-- active,ax ; Computed Time Active (ticks) cmp ax,14 ; Compare with 12 usec Specification ja DMA.sub.-- error ; If TA > 12usec, Then Issue Error jmp exiting ; Otherwise, Continue Normal OperationDMA.sub.-- error: mov .sub.-- TEXT : errors, OFFH ; Set Error Flag to Indicate the Error jmp exiting ; DRQ Not Active so Exitexiting:; pop ax ; Obtain Old 8259 IMR From Stack; out 021H,al ; Set up Old 8259 IMR cmp .sub.-- TEXT : CHN1.sub.-- flag,0 ; Is Channel 1 Disabled? jz not.sub.-- CHN1 ; If 0, Then not Disabled ; Otherwise, Must Be Enable mov al,01 ; DMA Channel 1 Enable Pattern out OAH,al ; Enable DMA Channel 1not.sub.-- CHN1: mov si,.sub.-- TEXT : SI.sub.-- reg ; Restore SI Register mov dx,.sub.-- TEXT : DX.sub.-- reg ; Restore DX Register mov cx,.sub.-- TEXT : CX.sub.-- reg ; Restore CX Register mov bx,.sub.-- TEXT : BX.sub.-- reg ; Restore BX Register mov ax,.sub.-- TEXT : AX.sub.-- reg ; Restore AX Registerchk.sub.-- old: cmp .sub.-- TEXT : ticks,128 ; Check to See if Old INT 8 is Needed jb send.sub.-- EOI ; If not Equal, Then Exit Timer ISR mov .sub.-- TEXT : ticks,0 ; Otherwise, Zero Total Tick Count jmp .sub.-- TEXT : int8.sub.-- ptr ; And Calling Old INT 8 Routinesend.sub.-- EOI: push ax ; Save AX Register mov al,020H ; 8259 End of Interrupt (EOI) out 020H,al ; Send EOI to 8259 pop ax ; Restore Ax Registertimer.sub.-- exit: iret ; Return from Interrupt.sub.-- Timer.sub.-- isr endp__________________________________________________________________________
The actual process is similar to the previous routines, however there is a counter that is initialized which represents the time required to perform the code presented above. If this counter expires (counts down to zero) and the transition has not occurred (DREQ to DACK) then the DMA operation is aborted and an error condition is returned which will cause the operation to be retried by the operating system. Furthermore, since the DMA operation was terminated then the adjacent sector is not in danger of being corrupted by the operation.
Thus, through software DMA shadowing, it is possible to determine when the last byte of the transfer is about to be transferred. Therefore, it is possible to disable any hardware and/or software resource that is present in the system that can cause the last data byte's transfer to be delayed. The use of software DMA shadowing accordingly allows system software and device drivers to be used to create a critical region about the last data byte transfer that will ensure that the byte is transferred correctly (and not delayed).
The invention described herein provides a complete software implementation of a device driver that is capable of detecting an undetectable data corruption problem without hardware redesign and/or internal modification to an existing FDC. Furthermore, the unique and innovative approach taken which consists of DMA shadowing and use of a software decoding network allows the implementation of the invention to be small (approximately 3.5 kilobytes in a PC/MS-DOS environment) and only marginally degrade performance (<10%) and only during floppy write operations.
The number of FDCs installed in computer systems today is well over 20 million. In order to solve this problem the vendors of such devices have very few alternatives, of which most are extremely costly. Therefore, a software-only solution to this problem is a significant advance in the computer industry. Moreover, the robustness of the design allows the system and method of the present invention to dynamically adjust to processor speeds that encompass the original IBM Personal Computers executing at 4.77 Mhz to the latest workstations that execute at well over 50 MHz.
The present invention may be embodied in other specific forms without departing from its spirit or essential characteristics. The described embodiments are to be considered in all respects only as illustrative, and not restrictive. The scope of the invention is, therefore, indicated by the appended claims, rather than by the foregoing description. All changes which come within the meaning and range of equivalency of the claims are to be embraced within their scope.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US3908099 *||Sep 27, 1974||Sep 23, 1975||Gte Automatic Electric Lab Inc||Fault detection system for a telephone exchange|
|US4942606 *||Feb 7, 1989||Jul 17, 1990||Compaq Computer Corporation||Computer with improved keyboard password functions|
|US4996690 *||Aug 26, 1988||Feb 26, 1991||Stac Electronics||Write operation with gating capability|
|US5093910 *||May 31, 1991||Mar 3, 1992||United Technologies Corporation||Serial data transmission between redundant channels|
|US5212795 *||Feb 10, 1992||May 18, 1993||California Institute Of Technology||Programmable DMA controller|
|US5233692 *||Jan 22, 1992||Aug 3, 1993||Micro Technology, Inc.||Enhanced interface permitting multiple-byte parallel transfers of control information and data on a small computer system interface (SCSI) communication bus and a mass storage system incorporating the enhanced interface|
|US5237567 *||Oct 31, 1990||Aug 17, 1993||Control Data Systems, Inc.||Processor communication bus|
|1||Adams, Phillip M., Nova University, Department of Computer Science, "Hardware-Induced Data Virus," Technical Report TR-881101-1 (U.S.A., Nov. 14, 1988).|
|2||*||Adams, Phillip M., Nova University, Department of Computer Science, Hardware Induced Data Virus, Technical Report TR 881101 1 (U.S.A., Nov. 14, 1988).|
|3||Advanced Military Computing, "Hardware Virus Threatens Databases," vol. 4, No. 25, pp. 1 & 8 (U.S.A., Dec. 5, 1988).|
|4||*||Advanced Military Computing, Hardware Virus Threatens Databases, vol. 4, No. 25, pp. 1 & 8 (U.S.A., Dec. 5, 1988).|
|5||Intel Corporation, "8237A/8237A-4/8237A-5 High Performance Programmable DMA Controller" (U.S.A., date unknown).|
|6||Intel Corporation, "8272A Single/Double Density Floppy Disk Controller" (U.S.A., Date unknown).|
|7||*||Intel Corporation, 8237A/8237A 4/8237A 5 High Performance Programmable DMA Controller (U.S.A., date unknown).|
|8||*||Intel Corporation, 8272A Single/Double Density Floppy Disk Controller (U.S.A., Date unknown).|
|9||*||Intel Corporation, Letter to customers from Jim Sleezer, Product Manager, regarding FDC error and possible solutions (U.S.A., May 2, 1988).|
|10||NEC Electronics Inc., "IBM-NEC Meeting for uPD765A/uPD72065 Problem" (U.S.A., May 1987).|
|11||*||NEC Electronics Inc., IBM NEC Meeting for uPD765A/uPD72065 Problem (U.S.A., May 1987).|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US5594865 *||Mar 25, 1996||Jan 14, 1997||Fujitsu Limited||Watchdog timer that can detect processor runaway while processor is accessing storage unit using data comparing unit to reset timer|
|US5842002 *||May 30, 1997||Nov 24, 1998||Quantum Leap Innovations, Inc.||Computer virus trap|
|US5862337 *||Jul 12, 1996||Jan 19, 1999||Microsoft Corporation||Determining throughput dynamically|
|US5983002 *||Oct 11, 1996||Nov 9, 1999||Phillip M. Adams & Associates, L.L.C.||Defective floppy diskette controller detection apparatus and method|
|US6195767 *||Sep 14, 1998||Feb 27, 2001||Phillip M. Adams||Data corruption detection apparatus and method|
|US6243833 *||Aug 26, 1998||Jun 5, 2001||International Business Machines Corporation||Apparatus and method for self generating error simulation test data from production code|
|US6279122||Dec 21, 2000||Aug 21, 2001||International Business Machines Corporation||Apparatus and method for self generating error simulation test data from production code|
|US6401222||Dec 14, 1998||Jun 4, 2002||Phillip M. Adams||Defective floppy diskette controller detection apparatus and method|
|US6687858||May 16, 2000||Feb 3, 2004||Phillip M. Adams||Software-hardware welding system|
|US6691181||Oct 9, 2001||Feb 10, 2004||Phillip M. Adams||Programmatic time-gap defect detection apparatus and method|
|US6751038||Jun 5, 2000||Jun 15, 2004||Kabushiki Kaisha Toshiba||Data write control system and method therefor|
|US6842802||Nov 30, 2001||Jan 11, 2005||Aftg-Tg, L.L.C.||Programmatic time-gap defect correction apparatus and method|
|US6889342 *||Jul 27, 2001||May 3, 2005||Hewlett-Packard Development Company, L.P.||Preventing floppy disk data corruption|
|US7069475||Nov 17, 2003||Jun 27, 2006||Aftg-Tg, L.L.C.||Software-hardware welding system|
|US7225361 *||Feb 21, 2002||May 29, 2007||Wily Technology, Inc.||Detecting a stalled routine|
|US7249203||Nov 17, 2003||Jul 24, 2007||Aftg-Tg, L.C.C.||Programmatic time-gap defect detection apparatus and method|
|US7251752 *||Oct 1, 2001||Jul 31, 2007||Adams Phillip M||Computerized product improvement apparatus and method|
|US7363237 *||Dec 4, 2003||Apr 22, 2008||Adams Phillip M||Enforcement process for correction of hardware and software defects|
|US7366804||Dec 10, 2004||Apr 29, 2008||Aftg-Tg, L.L.C.||Programmatic time-gap defect correction apparatus and method|
|US7409601||Jun 22, 2006||Aug 5, 2008||Phillip M. Adams & Associates, L.L.C.||Read-write function separation apparatus and method|
|US7472207||Jun 15, 2005||Dec 30, 2008||Aftg-Tg, L.L.C.||Optimized-incrementing, time-gap defect detection apparatus and method|
|US7653766||Apr 23, 2008||Jan 26, 2010||Adams Philip M||Time-gap defect detection apparatus and method|
|US7747452||Apr 21, 2008||Jun 29, 2010||Adams Phillip M||Enforcement process for correction of hardware and software defects|
|US8020030 *||Sep 18, 2008||Sep 13, 2011||Adams Phillip M||Counter-invasive software system and method|
|US8140910 *||Aug 4, 2008||Mar 20, 2012||Phillip M. Adams & Associates||Mismatched operation and control correction|
|US8645591||Jan 26, 2012||Feb 4, 2014||Aftg-Tg, Llc||Time-gap defect detection apparatus and method|
|US8667338||Feb 13, 2012||Mar 4, 2014||Aftg-Tg, Llc||Software-hardware read/write integrity verification system|
|US8732515 *||Aug 11, 2011||May 20, 2014||Phillip M. Adams||Counter-invasive software system and method|
|US9355013||Apr 15, 2014||May 31, 2016||Phillip M. Adams||Counter-invasive software system and method|
|US20020046367 *||Oct 15, 2001||Apr 18, 2002||Jin-Hsin Yang||Method for preventing data corruption by a floppy diskette controller|
|US20020170036 *||Feb 21, 2002||Nov 14, 2002||Cobb Jeffrey R.||Detecting a stalled routine|
|US20020184576 *||Mar 29, 2001||Dec 5, 2002||International Business Machines Corporation||Method and apparatus for isolating failing hardware in a PCI recoverable error|
|US20030023907 *||Jul 27, 2001||Jan 30, 2003||Neudeck Alexander J.||Preventing floppy disk data corruption|
|US20030065978 *||Oct 1, 2001||Apr 3, 2003||Adams Phillip M.||Computerized product improvement apparatus and method|
|US20030105897 *||Nov 30, 2001||Jun 5, 2003||Adams Phillip M.||Programmatic time-gap defect correction apparatus and method|
|US20040128675 *||Nov 17, 2003||Jul 1, 2004||Adams Phillip M.||Software-hardware welding system|
|US20040153797 *||Nov 17, 2003||Aug 5, 2004||Adams Phillip M.||Programmatic time-gap defect detection apparatus and method|
|US20050257101 *||Jun 15, 2005||Nov 17, 2005||Adams Phillip M||Optimized-incrementing, time-gap defect detection apparatus and method|
|US20050273528 *||Dec 10, 2004||Dec 8, 2005||Adams Phillip M||Programmatic time-gap defect correction apparatus and method|
|US20060242469 *||Jun 22, 2006||Oct 26, 2006||Adams Phillip M||Software-hardware welding system|
|US20120030763 *||Aug 11, 2011||Feb 2, 2012||Adams Phillip M||Counter-invasive software system and method|
|USRE44131||Jun 29, 2001||Apr 2, 2013||Fujitsu Limited||Storage device having function for coping with computer virus|
|U.S. Classification||714/55, 714/748, 714/749, G9B/20.046, G9B/20.051, 714/E11.003, 714/E11.062, G9B/27.052|
|International Classification||G06F11/16, G11B27/36, G06F11/00, G06F11/34, G11B20/18|
|Cooperative Classification||G06F11/0757, G11B27/36, G06F11/1612, G11B20/1816, G11B20/18, G06F11/3485|
|European Classification||G06F11/07P2A1, G11B20/18C, G06F11/16B2, G11B27/36, G11B20/18|
|May 23, 1995||CC||Certificate of correction|
|May 9, 1997||AS||Assignment|
Owner name: PHILLIP M. ADAMS & ASSOCIATES L.L.C., UTAH
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:ADAMS, PHILLIP M.;REEL/FRAME:008495/0085
Effective date: 19970422
|Jun 22, 1998||FPAY||Fee payment|
Year of fee payment: 4
|Mar 28, 2000||AS||Assignment|
Owner name: PACKARD, PACKARD & JOHNSON, CALIFORNIA
Free format text: ROYALTY & LIEN AGREEMENTS;ASSIGNOR:PHILIP M. ADAMS & ASSOCIATES, LLC;REEL/FRAME:010710/0575
Effective date: 20000327
|Jul 18, 2000||AS||Assignment|
Owner name: ADAMS AND ASSOCIATES, LLC, UTAH
Free format text: RELEASE AND ABANDONMENT OF INTERESTS;ASSIGNOR:PACKARD, PACKARD & JOHNSON, A PROFESSIONAL CORPORATION;REEL/FRAME:011170/0896
Effective date: 20000718
|Mar 27, 2002||FPAY||Fee payment|
Year of fee payment: 8
|Apr 12, 2006||FPAY||Fee payment|
Year of fee payment: 12
|Jul 8, 2010||AS||Assignment|
Owner name: SONY CORPORATION, JAPAN
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:PHILLIP M. ADAMS & ASSOCIATES. L.L.C.;REEL/FRAME:024651/0227
Effective date: 20100630